does max volume in mastering have more to do with compression and limiting in mixing?

offcenter2005

New member
My question is:

To achieve Professional mixed and mastered audio volume on your home recording tracks should you compress most or even all tracks in a project individually to a degree so it evens them out. I'm not talking about smashing their signals but just enough to even out transient peaks? Ive come close to matching volumes but the clarity is not there. I'm convinced that the problem is in the mix. Maybe the drums need compressed more before mix down or something along those lines. Any tips would be greatly appreciated. I'm using Izotope ozone 4 for mastering and have some waves plugins that i purchased off a friend who used to run a studio but gave up recording because of family life. That was a shame for him but a great thing for me.
 
Id also like to add that i record most of my instruments direct. I record direct guitars and use battery3 and ez drummer for drum tracks.
 
I generally find that whatever is done for the sake of volume will generally come back to haunt you later.

Far and away, the loudest mixes that come out of here are the most dynamic and "quietest" (a.k.a. "most normal") when they come in.

Do what serves the mix. Volume is the easy part.
 
Yeah, mixing is about perfecting the balance among your various tracks, not achieving a specific level. Compression can be a valuable tool in a mix but more for evening out levels than making things louder. For example, if you have one instrument or voice that has very little dynamic range and a second one that varies from very soft to very loud, you'll tend to loose the quiet bits on the track with a wide dynamic range. Compression, used carefully and subtly can even this out.

(For the sake of argument, I'm calling it all compression--even if sometimes the best way to achieve this is with "manual" compression, i.e. riding a fader!)

Beyond all that though, there's all sorts of psychoacoustics that can come into play at the mastering stage as you make things SEEM louder by creative use of compression. After all, 0dBFS is a brick wall you can't go through but compression to get everything up to that wall can make things seem louder than just a few peaks getting there. It also suck all the life out of your music, of course but that seems the trend in these days of most listening happening on earbuds in noisy places.

Bob
 
My question is:

To achieve Professional mixed and mastered audio volume on your home recording tracks should you compress most or even all tracks in a project individually to a degree so it evens them out. I'm not talking about smashing their signals but just enough to even out transient peaks? Ive come close to matching volumes but the clarity is not there. I'm convinced that the problem is in the mix.

If you're looking for clarity and definition in the mix, I think eq and level balance is the answer more so than compression.
 
ive also heard that you should not compress distorted guitars on input or in mix and if you do compress very very little . Is that good advice? And will that apply more so when recording direct than when recording miced?
 
(Aw geez...)

You compress what you need to compress when you need to compress it and how much it needs to be compressed to serve the dynamic content of the mix.

Dammit.

That said -- I don't think I've ever needed to compress a distorted guitar. Rarely a clean one for that matter. NEVER on the input... Back in the 70's when you had two exceptionally decent compressor and you needed to compress the input, it was one thing. Now, only if I need a certain flavor from a specific unit.
 
Look massive tude! Im trying to learn. I understand the song will tell me what needs done and not done but im asking technique questions so i can add ideas to my process. My songs sound good but im looking for ways to make them sound better. The only way to learn is to ask and apply. Distorted guitars are by nature already compressed, but maybe someone had a tip to get them sounding a little fuller without using up more headroom. Im sure when someone first said you could parallel compress the drums someone said (AW GEEZ!) On top of that modern mixes have no dynamics. Modern pop music for one. And by the way that is a pretty big musical market these days. If you feel that you are to good to explain some things to a guy who is willing to take advice and actually use it dont respond to my threads. If you feel like you still want to, keep the im better than this question shit for someone who dont know how to tune a guitar. I like posting with people who are willing to help. Thanks for your input but a little less (another newb question) would be fantastic.
 
The reason for some of the "roll eyes" responses is because the "max volume-mastering" question comes up weekly...just do a thread search.

I don't compress going in on hardly anything. I use to do some very light compression sometimes on bass or guitars (sometimes leads sometimes rhythm) but have gotten away from even that lately.
Occasionally I'll still compress my Hammond organ when tracking, mainly 'cuz certain voices tend to pop out while others become more subdued as you move up/down the keyboard...but it's still very light compression.

On mixdown...I compress my drum stereo OH tracks, a touch on bass, and honestly, that's about it, generally speaking, though each song is different. Mind you, I mix OTB using outboard hardware...not ITB with plugs.

I've been toying with the idea of adding some stereo bus compression just for flavor, not so much "max-volume", but I'm still waiting to buy a high-end hardware stereo comp/limiter, which is in the $2.5k to $3.5k range for my budget (there's even higher priced stuff, but too rich for my blood).
I end up doing most of my "compression" by hand, when editing in the DAW. I manually adjust peaks, and while it takes a lot more work and many passes to do that to multiple tracks, it's much tamer on the dynamics, EQ and balance than just slappin' a comp across the entire track (or stereo bus).

** Hey John - side question on level changes relating to EQ.

I've been aware of this "anomaly" for awhile...but right now that actual math behind it escapes me, and I think you might know how to explain it. When EQing the low end...there's this interesting level change that occurs.
What is the technical reason that when *cutting* LF by a certain amount on some broader bandwidth signal, that the overall level will actually *rise* a bit instead of falling?
I knew the answer...but I can't remember now. :D
 
thanks for that reply. I appreciate it greatly. I know there are more posts on this but they always tend to veer from the subject. When you manually adjust peaks do you just look for ones that are a lot higher than others on a specific track and isolate then do a gain reduction? And how would you determine which ones are too high? Sorry i know this is probably a redundant lot of questions but i need to learn this from experienced people because all of the guys in my area who record, mix and master are hacks that charge way to much and end up making the songs worse after they get a hold of a track or album. Not to talk bad about those guys but they just dont take time to learn this stuff and abuse their equipment instead of using it tastefully, Thank you again.

---------- Post added at 18:30 ---------- Previous post was at 18:25 ----------

sorry but another thin is im using some waves stuff and izotope ozone 4. I like everything on ozone but the loudness maximizer. would using the waves L2 be better for that instead?
 
I've been aware of this "anomaly" for awhile...but right now that actual math behind it escapes me, and I think you might know how to explain it. When EQing the low end...there's this interesting level change that occurs.
What is the technical reason that when *cutting* LF by a certain amount on some broader bandwidth signal, that the overall level will actually *rise* a bit instead of falling?
If you're talking about a signal that's "in the act" of being dynamically compromised, it's just that the low end takes 90% of the energy, but your ear is much more sensitive to the higher mids. Pull out the lows and the level bounces up because that energy is missing --- Less gain reduction / higher volume. Same thing happens with the ear to some extent --

sorry but another thin is im using some waves stuff and izotope ozone 4. I like everything on ozone but the loudness maximizer. would using the waves L2 be better for that instead?
Just sayin' -- Most people I know with Ozone use it for the loudness maximizer and don't like anything else.
 
When you manually adjust peaks do you just look for ones that are a lot higher than others on a specific track and isolate then do a gain reduction? And how would you determine which ones are too high?

It's part visual in the DAW and part listening and looking at the meter when the peak hits.
The "gain reduction" is done by cutting/cross-fading before and after each peak, and then manually pulling down its level.

Which peaks...?
Well, I start with the most obvious...and keep going from there, pass after pass for each track, until I get to the desired amount. By removing the offending peaks, I can then raise the level of the track --- manual compression.

---------- Post added at 19:15 ---------- Previous post was at 19:09 ----------

If you're talking about a signal that's "in the act" of being dynamically compromised, it's just that the low end takes 90% of the energy, but your ear is much more sensitive to the higher mids. Pull out the lows and the level bounces up because that energy is missing --- Less gain reduction / higher volume. Same thing happens with the ear to some extent --

No...not talking about using any compression, and not talking about what my ears are hearing.

In my DAW, if I mark off a lead guitar riff and then pull up an EQ plug and simply roll off the LF from say---150 Hz down, the metering will show a RISE in overall level of that guitar riff by a few dB compared to the pre-EQ metering level.
I've read somewhere about this effect...but can't recall why it causes the overall level to rise.
Logic would suggest that rolling "off" anything, would reduce some level overall.
Seems to only occur with the LF....
 
Just sayin' -- Most people I know with Ozone use it for the loudness maximizer and don't like anything else.

In comparison to the L2 it seems to me you can push the level a little harder before it starts distorting. Thats my opinion though, and since im new at mastering I could just be using it all wrong.
 
In my DAW, if I mark off a lead guitar riff and then pull up an EQ plug and simply roll off the LF from say---150 Hz down, the metering will show a RISE in overall level of that guitar riff by a few dB compared to the pre-EQ metering level.
I've read somewhere about this effect...but can't recall why it causes the overall level to rise.
Logic would suggest that rolling "off" anything, would reduce some level overall.
Seems to only occur with the LF....

There's a bump right above the shoulder frequency that occurs because of the phase shift in eq's that are minimal phase... so you will often see a increase of level when using a high pass filter.

Doesn't happen with linear phase eq's.
 
It's part visual in the DAW and part listening and looking at the meter when the peak hits.
The "gain reduction" is done by cutting/cross-fading before and after each peak, and then manually pulling down its level.

Which peaks...?
Well, I start with the most obvious...and keep going from there, pass after pass for each track, until I get to the desired amount. By removing the offending peaks, I can then raise the level of the track --- manual compression.

---------- Post added at 19:15 ---------- Previous post was at 19:09 ----------



No...not talking about using any compression, and not talking about what my ears are hearing.

In my DAW, if I mark off a lead guitar riff and then pull up an EQ plug and simply roll off the LF from say---150 Hz down, the metering will show a RISE in overall level of that guitar riff by a few dB compared to the pre-EQ metering level.
I've read somewhere about this effect...but can't recall why it causes the overall level to rise.
Logic would suggest that rolling "off" anything, would reduce some level overall.
Seems to only occur with the LF....

Im sure that a lot of my problem is in my mix i guess. I still dont have a complete grasp on frequency range and overall mixing in general but I spend about 8 hours a day trying to improve. I really like the new green day album for the sound of the guitars and it cranks. I am trying to separate the instruments in my ears to hear how they achieve the overall sound. The guitars seem thin and mid rangy but the low end of it comes from the bass im sure. I wish i could find a master track of each instrument after eqing and compression and all the other processing so i could get a good idea of how they do that end of it before the mastering. Maybe if id post a track i have of drums and guitar tracks you guys could give some pointers on what you think i should do to help me along. Thanks again.
 
There's a bump right above the shoulder frequency that occurs because of the phase shift in eq's that are minimal phase... so you will often see a increase of level when using a high pass filter.

Doesn't happen with linear phase eq's.

Ah...yes, that's what it was. I knew there was a legit/math reason...just forgot, and the other night I was rolling of the lows on some stuff, and again noticed that slight rise in level...so I wanted to remember why it did that.

My recollection is that linear phase EQs have their own set of "cons" (even though "linear" makes 'em sound more perfect)....
...but alas, I can't recall the reason for that, either! :D
 
No...not talking about using any compression, and not talking about what my ears are hearing.

In my DAW, if I mark off a lead guitar riff and then pull up an EQ plug and simply roll off the LF from say---150 Hz down, the metering will show a RISE in overall level of that guitar riff by a few dB compared to the pre-EQ metering level.
I've read somewhere about this effect...but can't recall why it causes the overall level to rise.
Logic would suggest that rolling "off" anything, would reduce some level overall.
Seems to only occur with the LF....

There will be places in the signal where a negative peak of one frequency lines up with a positive peak of another frequency so that the actual signal peak level is lowered at that spot. If you filter out one of the frequencies the other will peak higher.

Here's an example using DTMF (because it has more than one frequency at a time). First pic is a DTMF string at -12dBFS without filtering:

DTMF_-12dB_unfiltered.JPG

Second pic is the same file with a steep 1kHz HPF applied:

DTMF_-12dB_1kHz_HPF.JPG

The peak level of the whole file actually goes up 2.1dB, from -12dBFS to -9.9dBFS.
 
My recollection is that linear phase EQs have their own set of "cons" (even though "linear" makes 'em sound more perfect)....
...but alas, I can't recall the reason for that, either! :D
That's pre-ringing/echo that becomes more audibly apparent when used aggressively.
 
I hope this doesn't sound like im not understanding but the better the mix is to begin with then there will be less work to be done when mastering? A great mix and more attention to individual tracking will essentially lead to a louder overall sound when mastered? And that has to do with knowing where and when to compress, eq, limit and instrument placement in the stereo field? Is there any videos on the web i could check out to get a basic understanding of these factors? Please excuse my ignorance but i have a good ear when it comes to tracking and getting decent sounds but its just getting things to meld and sit where they should.
 
Back
Top